I know that some silicon logic ICs do emit light, but very little of it. The effect has been used for logic analysis, but needs tons of signal averaging
Bipolar transistors can emit light, too.
-- John Larkin Highland Technology Inc www.highlandtechnology.com jlarkin at highlandtechnology dot com Precision electronic instrumentation
No, I think the polarity is wrong there.
-- John Larkin Highland Technology Inc www.highlandtechnology.com jlarkin at highlandtechnology dot com Precision electronic instrumentation
The old CA3140 op amp sucked about 9 mA quiescent current, about a quarter watt. Heating doesn't happen because of the leakage currents, but because of the output drive stages.
Sure, static power dissipation warms up a chip, and should be minimized, but where is the positive feedback?
-- John Larkin Highland Technology Inc www.highlandtechnology.com jlarkin at highlandtechnology dot com Precision electronic instrumentation
Consider a transistor that sinks a fixed current, as part of the output drive. Even with zero load current, that transistor sinks 1 mA from +1V when the output state is above ground, and sinks 1 mA from -1V when the output state is below ground... So, the heat dumped at that part of the IC is modulated by the output voltage. If that part of the chip is near to the (+) protection diode on the (+) terminal, and distant from the other protection diodes, it will tend to pull the (+) terminal high when the output is high, by increasing the leakage.
The result (of the positive feedback) is that some operating conditions become bistable- a near-zero input leaves the output stuck in its old state.
As in N? Then just adjust substitute Vth with Vth' = N k_B T / q_e (er, or 1/N?).
Unless you mean N varies with V, I or T, which I've never heard before -- but, N always seemed uselessly dumb to me: if it is just a constant, you can fit the curve by Is just as well; SPICE models often do!
Or to put it another way, just measure it at a different temperature where Vth' is whatever intended value.
Tried looking in AoE but didn't find anything rmotely pertinent in the index. Why doesn't this thing have Google in it?...
Tim
-- Seven Transistor Labs Electrical Engineering Consultation Website: http://seventransistorlabs.com
"John Larkin" wrote in message news: snipped-for-privacy@4ax.com...
Bizarrely roundabout, but believable; crack open a 2N3055 (because why would you want to actually use one?) and reverse an ampere through the base; in a dim room, you'll be able to see the faint yellow glow. Be sure to check that you aren't measuring ambient light also.
Tim
-- Seven Transistor Labs Electrical Engineering Consultation Website: http://seventransistorlabs.com
Here is one that I made (for a moving coil meter):
You might be better off with a charge amplifier for this application, i.e. connect the non-inverting input of the op-amp to ground or a fixed voltage, and connect a polystyrene or teflon or polypropylene capacitor from the output of the op-amp to the inverting input of the op-amp, and feed the input current to the inverting input of the op-amp. Putting a small resistor (3k3 or so) in series with the input current may help with stability. You can reset the capacitor charge with a very clean reed relay, or with a circuit that uses PN4117 jfet devices as diodes that normally are configured to have no bias across the diode, but apply bias when you need to reset the capacitor.
Chris
On a sunny day (Sun, 01 Jun 2014 16:04:30 +1000) it happened Chris Jones wrote in :
Nice!
Pease suggested 2N2222 but other jelly bean parts work too. I've used
2N3904 and BC547 types. The few hundred millivolts negative in an otherwise positive only circuit can be used as a handy micropower source for op-amp offset null trim.piglet
To minimize leakage current from the traces of the input it might be a good idea to add guard traces around the input trace. I understand this is common in commercial devices.
-- Rick
Right. The so-called diode equation is ridiculously inaccurate in modelling real diodes, because what we laughingly call "high-level injection" effects dominate down to very low current densities. The equation works great for diode-connected transistors, at least for current densities where beta remains high. (A diode-connected transistor is the world's simplest negative feedback amplifier.)
No, it affects the logarithmic slope, not just the magnitude. exp(2x) is pretty different from exp(x). The forward current of a diode connected transistor goes up by a factor of 10 in about 60 mV, whereas your average diode diode requires 80-120 mV.
I lent my copy to my son, but you can probably find it in the list of figures.
Cheers
Phil Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 hobbs at electrooptical dot net http://electrooptical.net
Even if there were positive thermal feedback, you can get rid of it by using a compound amp made from two single op amps: run the first stage as a follower and the second stage as an integrator, with the feedback cap wrapped all the way round. That way the first stage's bias conditions (both input and output) are signal-independent, and the second stage's input current doesn't matter.
Cheers
Phil Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 hobbs at electrooptical dot net http://electrooptical.net
I've used an LT1028 followed by a fast unity-gain follower as a compound opamp. The 1028 needs really low-value feedback resistors to keep its noise down, and I didn't want my precision amp to source that much current and potentially make thermal tails. So the second stage does the heavy lifting.
-- John Larkin Highland Technology Inc www.highlandtechnology.com jlarkin at highlandtechnology dot com Precision electronic instrumentation
Reasonable. Of course the second stage has to be really fast, because the LT1028's frequency compensation is squirrelly to begin with, and you're talking about running it at lowish gains, I gather.
I've used the compound-integrator thing before in temperature control loops, where I wanted the first stage to be on the cold plate but didn't want any thermal forcing due to its bias conditions changing. Works great.
Cheers
Phil Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 hobbs at electrooptical dot net http://electrooptical.net
The 1028 has an "overcomp" pin that makes it all work. For the shunt amps in my NMR stuff, the first stage gain is typically 20, loop bandwidth 30 KHz maybe, so it wasn't hard to do. I did it by hand, pre-LT Spice!
-- John Larkin Highland Technology Inc www.highlandtechnology.com jlarkin at highlandtechnology dot com Precision electronic instrumentation
There is one of the Ti "electrometer" chips which has pins on one end set up for leakage islanding on the PCB. Pease put his IC's on 'stilts.' And about flux, he ran pcb through automatic dishwaher. Space wiring avoids a lot of problems, though. Bugs turned upside down over a pcb ground plane possible. jb
Hmmm Both Jim Williams and Bob Pease died at same time, roughly. I can't resist the urging to avoid that fate..."We get schmart too old." - Swedish saying.
On a sunny day (Sun, 01 Jun 2014 16:04:30 +1000) it happened Chris Jones wrote in :
Nice! PS this is a nice link talking about why the voltage increases when you move the capacitor plates away from each other:
Okay, that would make it less squirrelly, for sure.
Cheers
Phil Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 hobbs at electrooptical dot net http://electrooptical.net
something I made comment to years ago while talking about the math used in science.. A small group of us agreed that if you have to put in constants that were found by trial and error, Boltzmann, Planck etc.. then the subject matter was not yet fully understood.
It's like making up formulas using well know values and meanings in the subject matter but still does not resolve! So they throw in some made up constant! Sort of like winding coils, every one has their own math and usually includes a made up constant to make it work! Don't pay no attention to me, I am just disgruntle how many researchers just buy into it and say, well that's the way it is.
Why can't we just go back and fix some of the major errors in science so that we can move on. It wouldn't surprise me if there are two sets of science out there, one we are allowed to practice and not get to far, and the other, only the elite can know about it.
Jamie
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.